Claude Code and ScrapeGraphAI: lead scraping setup

This tutorial uses Claude Code to set up ScrapeGraphAI from a GitHub repository and request business leads in plain English. The example targets accountants, bookkeepers and tax services in the US, with the results exported to a CSV file. The source description identifies ScrapeGraphAI as open source.

The walkthrough starts with creating an OpenAI API key at platform.openai.com. The presenter then gives the repository link to the coding assistant and asks it to install the project on her computer. She uses the desktop version of Claude and follows the setup commands it supplies. In the configuration generated during her setup, a prompt asks for an Anthropic key and she enters an OpenAI key instead; the video does not establish this as the project’s standard installation method. A missing-payment message leads her to add funds to her OpenAI account. This setup uses a remote model API despite installing the scraping project locally.

The presenter requests 10,000 leads and reports receiving roughly 9,000 in less than an hour. She suggests allowing 30 to 90 minutes for a CSV to finish. These are her reported results and estimates, rather than guarantees for other searches; the walkthrough does not establish unlimited access or a universal limit for the project. The tutorial explains setup and export, but does not establish the accuracy or completeness of the collected leads.